On Fri, Nov 25, 2016 at 10:34:30AM +0000, Andrew Cooper wrote: > On 22/11/16 10:14, Jan Beulich wrote: > >>>> On 15.11.16 at 11:46, <jbeul...@suse.com> wrote: > >>>>> On 15.11.16 at 11:26, <andrew.coop...@citrix.com> wrote: > >>> Everything will indeed break if it is off by 8, but everything will also > >>> be similarly-broken if it is off by 16 or off by -8. > >> Off by -8 will be caught as well. And there's no possible off-by-16, > >> as then the calculations will be right again (read: you can freely > >> add two longs to struct cpu_info, but you mustn't add just one). > > This discussion appears to have stalled, and so far I've seen no > > reason to change other than the position of the checks. I'm > > hesitant to send out v2 though without having reached agreement > > on the other aspects. > > Fine - lets get this fix back in for 4.8, then argue over how it should > be improved. > > Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Release-acked-by: Wei Liu <wei.l...@citrix.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel